Impact of money raising considerations on campaign rhetoric

The virtue-based justification of electoralism implies an indirect connection between public opinion and policy. According to this justification, the public identifies people it trusts and puts them in office. Those people then determine policy as they see fit. According to this theory, then, the connection between policy and popular opinion is mediated by character judgments.

The rewards-based justification, on the other hand, implies a direct connection: elected officials wish to please the public in order to be re-elected and thus pursue policy that matches public opinion (in the sense that if they pursue policy X, then there exists no alternative policy which would win higher approval ratings).

The rewards-based theory suffers from two fundamental defects:

  1. It ignores the epistemic difficulties facing voters. In reality voters’ ability to determine the effects of government policy is very limited. They are therefore unable to tell whether government policy matches their world view and promotes their interests.
  2. It assumes that politicians lack policy preferences of their own. The theory assumes that politicians want to be elected simply and solely for pleasure of being in office rather than to promote any specific policy.

Those defects indicate why the rewards-based theory cannot be expected to explain policy setting by elected government. However, those defects do not apply to the rewards-based theory in the limited context of campaign rhetoric. While policy is difficult to analyze and has real effects in the world, campaign rhetoric is directly observable by the voters and has little effect other than as a means for gaining electoral advantage. Therefore, epistemic difficulties are by definition non-existent and politicians should have little reason to have inherent attachment to particular rhetoric, leaving them free to fashion their campaign rhetoric so as to maximize electoral rewards.

However, while it is true that campaign rhetoric hews to public opinion much more closely than policy, it is obviously not the case that campaign rhetoric matches public opinion. For example, the American public considers “jobs and unemployment” to be the most important issue facing the country. Yet, no candidate in the U.S. offers concrete proposals for increasing employment, although it is quite simple to imagine various strategies to do so. Other clear examples of opinions held by the U.S. public over long periods but not getting reflected in campaign rhetoric relate to taxation of corporations and the rich, campaign finance, and nuclear disarmament. One recent relevant case is Obama’s changing line on income inequality, where Obama appeared for a short while to be picking up a popular theme for his rhetoric, only to drop it soon afterwards.

It is interesting to consider, then, what explains the failure of the rewards-based theory to explain electoral behavior even in the limited context of campaign rhetoric.

One possible explanation is by the need to raise campaign funds. Obviously, the effect of campaign rhetorical in garnering votes depends not only on the content of the message being broadcast but also on the reach of the message (the number of people being exposed to the message and the frequency of exposure). In the U.S. system, the reach of campaign messages depends crucially on the campaign funds raised by the candidates. To be able to reach voters and potentially win their votes, candidates must win the backing of donors. Since potential donors are a different population than voters (especially when weighted by potential donation), campaign messages that resonate with voters may not resonate with donors, and vice versa. To optimize the number of votes, a candidate, therefore, must broadcast a message that will balance the characteristics of the two audiences: donors and voters.

This model can be formalized as follows:

v = v(r, f),

where v is votes won, r is rhetoric used and f is campaign funds spent. If funds available are fixed, then the candidate is free to choose r so as to maximize votes. However, the funds available depend themselves on the rhetoric used:

f = f(r).


v(r) = v(r, f(r)).

Thus, the candidate’s rhetoric optimizes over both pleasing voters and pleasing donors.

A simple concrete example is as follows:

v(r, f) = V rα fβ,

f(r) = 1 – r.

In this model, the responsiveness of the voters to the rhetorical content is characterized by α and the responsiveness to the quantity of exposure (equivalent to campaign funding) is β. The optimal rhetoric is

r* = α / (α + β).

That is, when the voters respond mostly to rhetorical content, campaign rhetoric will hew more closely to popular opinion, while relative responsiveness to exposure will shift the rhetoric to reflect donor opinion.

24 Responses

  1. If political behaviour could be reduced to algebraic equations and the simple binaries beloved of software engineers (virtue OR interests) then life would be a lot more straightforward. It would also be wonderful if simple mechanisms to increase (domestic) employment (without creating significant externalities) could be “imagined” in the world of global capital. In fact we could give up on politics completely and establish a cybernetocracy — as I’m sure a machine would do a better job of imagining the necessary strategies to reduce unemployment than fallible human agents. It would also save a lot of money (and provide employment opportunities for software engineers and others who like to imagine that these problems are simple to resolve). We could also do away with fusty old disciplines like political science, history, sociology, psychology and other messy attempts to understand human behaviour in its analogue complexity.


  2. PS by “externalities” I was referring to issues like inflation, budget deficits, public-private balance, global competitiveness etc. Political decision-making involves complex trade-offs (and that’s before we even get to to pork trading) and cannot be reduced to single vectors such as “interests”. If you insist on turning that into algebra, then the complexity involved would soon make it an entirely counter-productive endeavour. The algebraic/deductive approach you adopt may make sense in your own discipline, but is of no relevance to the study of human agents.


  3. We’ve had an exchange on the usefulness of formalization here. Since, as is your habit, you have added nothing to what you wrote over two years ago, I think re-reading the exchange we had then should suffice as a reply.


  4. Yoram,

    Your skills as an archivist are without parallel. Personally, I’m not in the slightest bit interested in what any of us might have said two years ago, as I would like to think that we are all capable of learning from each other and moving forward. Naomi’s recent contributions have caused me to revise my views considerably. How about you — has any of the exchange on this forum encouraged you to revise your own perspective on sortition? If so, then how?


  5. > I’m not in the slightest bit interested in what any of us might have said two years ago

    I find this quite credible as you habitually show essentially no interest in what anyone says, no matter how recently.


  6. I read all the comments very carefully, and acknowledge when they force me to alter my viewpoint; see for example my recent exchange with Naomi, or are your archival interests limited to your own posts?


  7. I was not aware of Bernard Manin, Adam Przeworski, and Susan C. Stokes (1999), which seems very relevant.


  8. Formal model or not, the fact that Congressional reps spend around 80% of their TIME raising funds should make one reconsider elections, not only as currently conducted in the US but elsewhere.

    What you spend your time doing is what you are. In this case, “people’s representatives” become public relations self-reps.


  9. Ahmed,
    That’s what happens when you have:

    1) Elections every two years.

    2) Weak parties. Candidates have to make name for themselves and not rely on the party’s existing brand so much.

    3) No public election financing.

    I have no idea how much time the average (for example) German MP spends on fundraising but I’m willing to bet it’s not enough to be a problem.


  10. Ditto in the UK.


  11. Naomi,

    First, as a matter of fact, party “brand” is crucial in the US. There are very few candidates who manage to get elected without party affiliation.

    But more substantively, to the extent that the parameters you are referring to make any difference (and they probably make little difference), they may shift power from donors to the party apparatus. There is no reason to expect them to distribute power to the public at large.


  12. Of course belonging to one of the top two parties is essential. People don’t want to split their vote. Thanks to party primaries, the US has an approximation of a two-round majoritarian system, but only if you only have two candidates in the general election. Almost anyone who could win a non-partisan general election could win one of the primaries. This, combined with the presidential coattails effect, makes minor parties (and independent candidacies) vastly more difficult to sustain. The interparty dynamic is squashed in favor of the intraparty dynamic.

    >they may shift power from donors to the party apparatus.

    Good. This should be a goal. It is best to think of the parties themselves as the candidates in party centric systems. The individual MPs matter little. I think we can all agree that shifting power from donors to the candidates is good.


  13. > I think we can all agree that shifting power from donors to the candidates is good.

    Funny – I would think we could all agree that shifting power from donors to the party apparatus (or vice versa) is meaningless, since both those groups represent elite interests rather than the public interest.


  14. This one-dimensional focus on “elites” as the root of all evil is wearing a bit thin, don’t you think?


  15. Yes – a bit tired repeating the same claims over and over simply because they appear to be valid.


  16. Your claim may well be valid in terms of your deductive formalism, but the real world is multi-dimensional and analogue in nature and not reducible to simple algebraic equations with binary outcomes (elite-v-public; virtue-v-interests). The motivations underlying political behaviour are complex and not open to demonstration by propositional calculus.


  17. Yoram,
    I have to admit, even after reading through a good chunk of the archives, I’m not 100% sure what your focus on the “elites” is about. Something about confusing and distracting the masses from acting in their common interest… which results in a reduced standard of living for most people? Or something along those lines?


  18. Keith,
    For what it’s worth, I don’t necessarily think the basic premise of the article above is completely wrong. You do need money to get your message out, which is why there’s so much pressure on US House candidates to fundraise. I just think it’s a solved problem.


  19. Naomi,

    I don’t deny the existence of the problem, only the methodology employed (deductive formalism), which is of little value in empirical disciplines such as political science. On the rare occasions that it generates testable hypotheses (such as the frequently trumpeted Gilens paper), the statistical methodology and interpretation of results is flawed, as has been demonstrated at length on this forum. For example Gilens concludes that the median voter has no influence at all on policy outcomes, but the data merely indicates that in the minority of cases where median and high-income preferences diverge, there is a (marginally) closer link between outcomes and high-income preferences. And there is no attempt to analyse which issues median voters consider to be salient. I guess one of the differences between political science and (say) chemistry is that in the former example “experimenters” tend also to be political partisans (take a look at the Gilens/Page media coverage to see what I mean).


  20. Naomi,

    > Something about confusing and distracting the masses from acting in their common interest…

    I can’t recall writing anything like this (let alone it being my “focus”). Care to point me at a post where I wrote this?


  21. It was just the gist I got from reading through the archives. As I said, I don’t understand where you’re coming from.


  22. Naomi,

    I think most people would have a similar impression. In order to fully understand Yoram’s position you need to read the classic texts in political sociology — Marx, Mosca, Pareto, Michels, Mills — (whether these are of any relevance to modern pluralist democracies is a good question). You would also need to subscribe to the materialist worldview that interprets everything in terms of ” interests”, and views ideology as an epiphenomenon.


  23. > It was just the gist I got from reading through the archives.

    I think I may be able to understand now how you and Sutherland are getting along so well.

    > I don’t understand where you’re coming from.

    If you are interested, you can have a look at my introductory presentation: Sortition is natural to democracy, as elections are to aristocracy.

    If you have more specific questions, we can discuss those.


  24. >Sortition is natural to democracy, as elections are to aristocracy.

    Also Sprach Zarathustra. Naomi and I are more interested in the modern world than the very different conditions of a small polis 2,500 years ago. There is a lot that we can learn from Athenian experience, but this involves a bit more than just parroting Aristotle and Montesquieu.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: